WNT pathway inhibition sensitizes HAT1-high lung cancers to treatment with PD-1 inhibitors

Abstract

Background: Immunotherapies change the paradigm of current pulmonary oncological clinics, although majority of patients fail to benefit from these treatment modalities. HAT1 overexpression is frequently diagnosed in lung cancer patients. Effective immunotherapeutic scheme remains to be determined for this portion of patients.

Methods : In vitro experiments were used to verify the transcriptional regulation of PD-L1 by HAT1 in lung cancer cells. Mouse syngeneic lung tumors with sgHAT1 were treated with PD-1 antibodies to determine the impact of deficiency of HAT1 on the sensitivity of lung cancers to PD-1 inhibitors. RNA-seq and Chip-seq analyses revealed the effect of HAT1 on the WNT pathway. The impact of HAT1 on WNT signaling on stemness of lung cancer cells and their ability to escape immune surveillance were evaluated through a series of in vitro and in vivo experiments. Clinical relevance of expression level of HAT1 on prognosis of lung cancer patients were assessed using data from multiple public datasets. Carcinogenicity of HAT1 was evaluated using a transgenic mouse model in vivo.

Results: Silencing HAT1 expression downregulated PD-L1 expression in murine and human lung cancer cells. Notably, HAT1-knockout sensitized lung cancers to PD-1 antibody treatment. We found that HAT1 enhanced the stemness of lung cancer cells through enhancing expression of WNT ligands by direct binding of their promoter regions. WNT pathway inhibitor, ICG001, sensitized lung tumors to treatment with PD-1 antibody. HAT1 overexpression was highly frequent among lung cancer patients and was found transforming in vitro and tumorigenic in vivo.

Conclusion: This study demonstrates that overexpression HAT1 plays a critical role in mediating immune evasion in lung cancer. Mechanistically, on one hand, HAT1 upregulate PD-L1 expression in lung cancer cells. On the other hand, HAT1 enhances stem-like properties of lung cancer cells by upregulating several WNT ligands. Consequently, combined inhibition of the WNT pathway and PD-1 signaling effectively shrinks HAT1-high lung cancers. These findings highlight the critical role of HAT1 overexpression in lung cancer development and shed light on potential precision medicine for this portion of lung cancer patients.

Supplementary Information: The online version contains supplementary material available at 10.1186/s12935-025-04007-2.

Keywords: HAT1; Immunotherapy; Lung cancer; PD-1; WNT signaling.

Fig. 1

Knockout of HAT1 renders lung cancer tumors more sensitive to PD-1 antibody treatment than knockout of PD-L1. A. Establishment of TC1-sgHAT1 and TC1-sgPD-L1 cell lines. TC1 cells were infected with lentiv2-sgHAT1 lentivirus or lentiv2-sgPD-L1 lentivirus (Non-targeting (NT) sgRNA lentivirus serving as control) for 24 h, followed by selection with puromycin (1 µg/mL) for 7 days. Whole cell lysate was analyzed through immunoblot with indicated antibodies B. Schematics of treatment schedules for TC1 tumor C-E. TC1-sgHAT1 syngeneic tumors are more sensitive to αPD-1 treatment in vivo. 2 × 10⁶ TC1-sgHAT1 and 1.5 × 10⁶ TC1-sgPD-L1 cells (1 × 10⁶ TC1-sgNT as control) were inoculated subcutaneously in B6 mice (n = 8). Mice were randomized for treatment with saline (serving as Vehicle) or PD-1 antibody (n = 4). When syngeneic tumor reached about 80 mm³ (C). Tumors were dissected for photographing (D), weighing (E) by the end of experiments F. Establishment of LLC-sgHAT1 and LLC-sgPD-L1 cell lines. LLC cells were infected with lentiv2-sgHAT1 lentivirus or lentiv2-sgPD-L1 lentivirus (Non-targeting (NT) sgRNA lentivirus serving as control) for 24 h, followed by selection with puromycin (1 µg/mL) for 7 days. Whole cell lysate was analyzed through immunoblot with indicated antibodies G. Schematics of treatment schedules for LLC tumor H-J. LLC-sgHAT1 syngeneic tumors are more sensitive to αPD-1 treatment in vivo. 2 × 10⁶ LLC-sgHAT1 and 1.5 × 10⁶ LLC-sgPD-L1 cells (1 × 10⁶ LLC-sgNT as control) were inoculated subcutaneously in B6 mice (n = 8). Mice were randomized for treatment with saline (serving as vehicle) or PD-1 antibody (n = 4) when tumors reached about 80 mm³ (H). Tumors were dissected for photographing (I), weighing (J) by the end of experiments K-L. PD-1 inhibitor treatment elicits tumor cell apoptosis. Representative images of TC1 (K) or LLC (M) tumor sections were shown (Scale bars:200 μm for H&E and 100 μm for cleaved caspase3). Quantitative statistics represent Cleaved caspase3 positive cells (brown) of TC1 (L) or LLC (N) tumor sections. Bars are represented as means ± SD of three independent experiments, *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001

Fig. 2

HAT1-knockout lung tumors undergoes drastic rewiring of T cells in response to PD-1 antibody treatmentA. Gating strategy for analyzing CD4⁺ T and CD8⁺ T cells in tumor nodules B-E. T cell infiltration in tumors. TC1 or LLC tumor nodules were dissected for evaluating the population disparity of CD45⁺ CD4⁺ or CD45⁺ CD8⁺ T cells by flow cytometry. Statistics representing the percentages of CD4⁺ or CD8⁺ T cells in four different treated groups of TC1 tumor nodules (B-C) or LLC tumor nodules (D-E) F-I. αPD-1 treatment enhances production of cytokines by CD8⁺ cytotoxic T cell of TC1-sgHAT1 tumor nodules. Representative flow cytometry plots (F) and quantitative statistics showing the percentage of IFNγ (G), IL-2 (H) and Perforin (I) produced in tumor-infiltrating CD8⁺ T cells. The IFNγ⁺, IL-2⁺ and Perforin⁺ population were defined according to isotype-matched antibody staining control J-M. αPD-1 treatment enhances production of cytokines by CD8⁺ cytotoxic T cell of LLC-sgHAT1 tumor nodules. Representative flow cytometry plots (J) and quantitative statistics showed the percentage of IFNγ (K), IL-2 (L) and Perforin (M) produced in tumor-infiltrating CD8⁺ T cells. The IFNγ⁺, IL-2⁺ and Perforin⁺ population were defined according to isotype-matched antibody staining control. Bars are represented as means ± SD of three independent experiments, *P < 0.05, **P < 0.01 and ***P < 0.001

Fig. 3

HAT1 activates WNT/β-Catenin signaling pathway.A. KEGG analysis of downregulated genes in PC9 cell in response to HAT1 knockdown. A. KEGG analysis of downregulated genes in PC9 cell in response to HAT1 knockdown. PC9-KD cells were treated or untreated with DOX (0.5 µg/mL) for 48 h before RNA sequencing. Significantly downregulated genes were subjected to KEGG pathway enrichment B. Impact of HAT1 expression on Topflash reporter in PC9 cells. PC9-sh/+Vec and PC9-sh/+HAT1 cells were transfected with the Topflash reporter, followed by treatment with DOX (0.5 µg/mL) and LiCl (10 µM) for another 48 h before luciferase were monitored C. RT-qPCR analysis of impact of HAT1 expression on WNT/β-catenin target genes in PC9 cells. PC9-sh/+Vec and PC9-sh/+HAT1 cells was treated with DOX (0.5 µg/mL) for 48 h. Cells were harvested to evaluate C-MYC, TP63, CCDN1, CD44 and AXIN2 expression through RT-qPCR analysis D. Characterization of PC9i cells. PC9i cells were treated with/without DOX (0.5 µg/mL) for another 48 h. Whole cell lysate was analyzed through Immunoblot with the indicated antibodies E. Distribution of enrichment signal for all annotated genes of TSS and TES metagene profile of ChIP-Seq data of PC9i cells. Peaks of input sample (upper panel) and Flag antibody ChIP sample (lower panel) were shown for the entire genic area of annotated genes. The metagene coordinates on the x-axis show the genomic distance (in bp) from the TSS with up- and down-stream regions in real scale. Y-axis shows normalized enrichment signal in log² value F. IGV browser views of enrichment areas throughout WNT4 (Peak1: chr1 22,462,800 bp- 22,462,903 bp), WNT5B (Peak1: chr12 1,751,455 bp- 1,751,565 bp), WNT6 (Peak1: chr2 219,724,400 bp-219,724,537 bp) and WNT7B genes (Peak1: chr22 46,321,152 bp-46,321,355 bp) in PC9i cells G. HAT1 binds promoter regions of WNT4, WNT5B, WNT6 and WNT7B genes. PC9i cells were treated with DOX (0.5 µg/mL) for 48 h. ChIP-PCR was conducted to amplify WNT4 (103 bp), WNT5B (110 bp), WNT6 (137 bp) and WNT7B (203 bp) promoter DNAs H. RT-PCR quantification of mRNA expression of WNT4, WNT5B, WNT6 and WNT7B in PC9-sh/+Vec and PC9-sh/+HAT1 cells. Bars are represented as means ± SD of three independent experiments, *P < 0.05, **P < 0.01 and ***P < 0.001

Fig. 4

HAT1 regulates stemness of lung cancer cells. A-B. HAT1 knockdown reduces sphere formation capacity. PC9-shHAT1 (A) and H460-shHAT1 (B) cells (1,000 cells/well) were cultured in ultra-low attachment 6-well plates with or without DOX (0.5 µg/mL) for 14 days. Representative sphere images (upper panel) and quantitative analysis of sphere numbers (lower panel) are shown (Scale bar: 500 μm) C-D. HAT1 knockdown impairs anchorage-independent growth in soft agar. PC9-shHAT1 (C) and H460-shHAT1 (D) cells (1,000 cells/well) were subjected to soft agar assays with DOX (0.5 µg/mL) treatment for 14 days. Representative colony images (upper panels) and colony counts (lower panels) are presented (Scale bar: 500 μm) E-F. Percentage of CD44⁺ population decreases in response to HAT1 knockdown in PC9 (E) and H460 cells (F). Representative images of flow cytometry analysis (left) and statistics of CD44 MFI (right) G-H. HAT1 Knockdown downregulates frequency of side population cells in PC9 (G)and H460 cells (H). Side Population was analyzed through uptake of Hoechst33342 red in the presence of verapamil or not. Representative images of flow cytometry analysis (left) and statistics of stem cells (right) I. Ectopic HAT1 expression rescues the sphere-forming ability of PC9 cells. Representative images of spheres (upper panel) and statistics of colony numbers (lower panel) (Scale bar: 500 μm) J. Ectopic HAT1 expression rescues the 3D colony forming ability of PC9 cells. Representative images of soft agar culture (upper panel) and statistics of colony numbers of cells (lower panel) (Scale bar: 500 μm) K. Ectopic HAT1 expression upregulates the percentage of CD44⁺ subpopulation of PC9 cells. Representative images of flow cytometry analysis (left) and statistics of CD44 MFI (right) L. Ectopic HAT1 expression upregulates the percentage of side population of PC9 cells. Representative images of flow cytometry analysis (left) and statistics of stem cells (right). Bars are represented as means ± SD of three independent experiments, *P < 0.05, **P < 0.01 and ***P < 0.001

Fig. 5

The Combination ICG001 and αPD-1 effectively shrinks HAT1 overepxressing lung cancers. A-C. Treatment effect of ICG001 and αPD-1 on LLC allograft tumors. 1 × 10⁶ LLC cells were subcutaneously inoculated in right flank of B6 mice. Mice were randomized for treatment with saline (Vehicle), PD-1 antibody, ICG001 and combination of PD-1 antibody and ICG001, respectively (n = 4) when tumors reached a volume of about 80 mm³. Tumors volumes were recorded every 3 days (A). LLC tumors were dissected for photographing (B) and weighing (C) D-E. Statistics of tumor-infiltrating CD4⁺ T cells (D) or CD8⁺ T cells (E) in LLC tumor nodules revealed through flow cytometry F. Combination of αPD-1 and ICG001 treatment ameliorates exhaustion of CD8 cytotoxic T cells. Representative flow cytometry plots of PD-1, TIM-3 and LAG3 expression in LLC tumor-infiltrating CD8⁺ T cells G-I. Statistics of expression of TIM3, PD-1and LAG3 of CD8⁺ T cells in (F) J. Combination of αPD-1 and ICG001 enhances IFNγ, IL-2 and Perforin production of CD8⁺ cytotoxic T cells. Representative flow cytometry plots of IFNγ, IL-2 and Perforin expression in LLC tumor-infiltrating CD8⁺ T cells K-M Statistics of expression of IFNγ, IL-2 and Perforin of CD8⁺ T cells in (J). Bars are represented as means ± SD of three independent experiments, *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001

Fig. 6

HAT1 overexpression is tumorigenic and associated with poor prognosis in LUAD patients A. HAT1 mRNA expression in lung cancer tissues revealed in TCGA database and GTEX database B. Expression of HAT1 protein in LUAD tissues and normal lung tissues revealed in the UALCAN databases (http://ualcan.path.uab.edu) C. Comparison of expression of HAT1 protein in normal versus LUAD tissues in HPA database (https://www.proteinatlas.org/). Representative images for IHC staining of HAT1 in normal lung and LUAD tissues (left) (Scale bar: 200 μm). The box diagram of the HAT1 expression on IHC score (right) (normal lung tissues, n = 4; LUAD tissues, n = 15) D-E. Survival analysis of HAT1 expression in patients with LUAD cancer. The overall survival (OS) of patients with LUAD cancer derived from analysis of PRECOG Database (https://precog.stanford.edu/) (D) and the GEPIA database (http://gepia.cancer-pku.cn) (E) F. Overexpression of HAT1 is transforming. 3T3-Vec (4 × 10⁶) and 3T3-HAT1-oE cells (4 × 10⁶) were subcutaneously inoculated into right flank of nude mice (n = 4). Tumor-bearing mice were photographed at day 35 after inoculation (upper panel). Tumors were harvested and photographed by the end of experiments. (lower panel) G. HAT1 overexpression is tumorigenic in vivo. SPC-Cre-ER^T2; p53^fl/fl transgenic mice were treated with tamoxifen (2 mg/mouse) once a day, three times in total. After 2 days of rest, mice were infected with HAT1 overexpressing virus through nasal instillation. Mice were imaged through computed tomography (CT) scanning every month after virus infection. Representative CT images (right) of lentivirus infected SPC-Cre-ER^T2; p53^fl/fl transgenic mice (left) and quantitative statistics tumor burden (right) H. Representative photographs of lungs of mice 4 months after lentivirus infection I. Ki-67 staining of lung section of SPC-Cre-ER^T2; p53^fl/fl transgenic mice (brown signal) (left) and statistics of Ki-67 positive cells in tumor tissue (Scale bar: 500 μm and 100 μm) J. Hematoxylin and eosin (H&E) staining of lung section of lentivirus infected SPC-Cre-ER^T2; p53^fl/fl transgenic mice (left) and quantitative statistics of average tumor numbers (middle) and relative tumor area (right) lung sections (Scale bar: 500 μm and 100 μm) K. HAT1 overexpression drives formation of lung adenocarcinoma. Lung tumors from SPC-Cre-ER^T2; p53^fl/fl transgenic mice were subjected to IHC staining for expression of p63 (left) and TTF-1(right) (Scale bar: 500 μm and 100 μm). Bars are represented as means ± SD of three independent experiments, *P < 0.05, **P < 0.01 and ***P < 0.001